Make IABookreader respect the IIIF manifest

[DiegoPino]

What is this?

This comes from a google group discussion about making a IIIF manifest that works for a single PDF by using cantaloupe to extract each Page as individual images. See this demo Twig template https://gist.github.com/DiegoPino/724ce3b703f870acb63c3340c7f9d897

When enabling that Twig Templates as a Metadata Display in Archipelago and setting IABookreader to use it, and also assuming you have a PDF ingested in a given ADO, and you also manually added an extra key named "pageCount":7 (e.g) like here, which is needed until beta3 is out..

 "as:document": {
        "urn:uuid:64e52047-0740-41a0-9034-89c244558986": {
            "url": "s3:\/\/827\/application-14388-66424-1-pb-64e52047-0740-41a0-9034-89c244558986.pdf",
            "name": "14388-66424-1-PB.pdf",
            "tags": [],
            "type": "Document",
            "dr:fid": 6,
            "dr:for": "documents",
            "dr:uuid": "64e52047-0740-41a0-9034-89c244558986",
            "checksum": "827187fd60b0d77aa6127fbab4aed663",
            "sequence": 1,
            "pageCount": 7,
            "crypHashFunc": "md5"
        }
    },

You will get same page displayed, in my case 7 times.

Which is strange since the IIIF Manifest generates correctly Resource entries like this one

"resource":{
      "@id": "http://localhost:8183/iiif/2/827%2Fapplication-14388-66424-1-pb-64e52047-0740-41a0-9034-89c244558986.pdf/full/full/0/default.jpg?page=2",

with page varying from 1 to 7 in my case.

So, error is not the IIIF Manifest, but our/my/IA assumption on the JS on how to handle IIIIF V2.

In particular this line

https://github.com/esmero/format_strawberryfield/blob/8.x-1.0-beta3/js/plugin.iiif-iabookreader_strawberry.js#L106

Where instead of trusting the @id in the resource section, we try to be "smarter" and we recreate the call based on the Service ID... which is so good and so bad (good if we have like just images all with actual services, and info.json, etc, but bad for any one writing Manifests manually or needing to use PDF extraction like ourselfs.

Solution. Rewrite that JS to deal with rebuilding the the URL only, and only if, the ID is not already and image. Just sayng

@giancarlobi what do you think? Is my explanation clear?
@mitchellkeaney this is why you see your page twice..

[giancarlobi]

@DiegoPino , you are clear. I have to check that to fully understand it with some examples. What about next week? If some time free i'll check it before. Thanks!!!

[DiegoPino]

@giancarlobi is Monday good? =) Thanks

[giancarlobi]

@giancarlobi is Monday good? =) Thanks

Perfect!!

[giancarlobi]

@DiegoPino I'm ready to work on this, I was able to reproduce the issue http://archipelago.byterfly.eu/node/29.

[giancarlobi]

@DiegoPino A first (simple/stupid?) idea.
Manifest of a standard book (TIFFs) has into each Service ID a distinct URL to different images. Manifest of PDF has into each Service ID the SAME URL.
So, what about include "?page=x" into Service ID URL (by Twig template) and into code highlighted we cut and paste (if present) "?page=x" at the end of uri: rewrited?

[DiegoPino]

Hi, can you meeting in 45 min? Can be less too. Issue i see with our iabookreader js plugin is that it will also fail if the manifest has no service. I have seen many around where the images are served statically and not via a iiif server. Let me find some examples and discuss your idea. You mean adding the page argument to the service id?

[DiegoPino]

Sorry, yes you said service ID! (My fault) i agree, that would be the change on the twig template. So code needs yo have some extra work, need to check if cantaloupe returns an info.json that really changes by appending the ?page at the end

[giancarlobi]

need to check if cantaloupe returns an info.json that really changes by appending the ?page at the end

@DiegoPino ok at my 15:30 and cantaloupe doesn't return error appending ?page and I think it returns right json, to check. Meet you soon.

[giancarlobi]

@DiegoPino To talk: maybe here

    function getImageProperties(image) {
        var imageObj = {
            height:       image.height || 0,
            width:        image.width || 0,
            imageUrl:     image.service['@id'].replace(/\/$/, ''),
        };

we have to add another option imageQParam and cut QParam from imageUrl option?

[giancarlobi]

Then here
uri: image.imageUrl + "/full/" + image.width + ",/0/default.jpg" + "?" + image.imageQParam,
Obviously for standard tiff image.imageQParam will be NULL.